The deployment pipeline that broke in prod (and how I fixed it)
A green build, a broken release, and the missing piece between them.
Tools matter less than the feedback loop they enable. A fast, boring loop beat every sophisticated setup I tried, and I've kept it ever since, even as the rest of the stack changed twice over.
Setting the scene
There is a version of this article that opens with theory. I decided against it, because the theory is easy to find and the motivation is not — so let's start with the thing that actually broke.
There is a version of this article that opens with theory. I decided against it, because the theory is easy to find and the motivation is not — so let's start with the thing that actually broke.
The naive approach
What surprised me most was how small the final change was. Most of the work was understanding the problem precisely enough that the fix became obvious, and that understanding took far longer than the fix itself.
What surprised me most was how small the final change was. Most of the work was understanding the problem precisely enough that the fix became obvious, and that understanding took far longer than the fix itself.
A note for context that helps the reader.
const debounce = (fn, ms = 200) => {
let t;
return (...args) => {
clearTimeout(t);
t = setTimeout(() => fn(...args), ms);
};
};What the profiler said
The first version worked, which is exactly why it was hard to see how wrong it was. Everything looked fine in the demo, and everything looked fine until traffic arrived on a Tuesday afternoon and the graph started doing something I'd never seen before.
Tools matter less than the feedback loop they enable. A fast, boring loop beat every sophisticated setup I tried, and I've kept it ever since, even as the rest of the stack changed twice over.
The best engineering advice I've received is unglamorous: make it measurable, make it reversible, and only then make it fast.
The rewrite
By the end of this post you should be able to reproduce the whole thing yourself in under an hour, including the parts I got wrong the first time — those are marked so you can skip them if you're in a hurry.
What surprised me most was how small the final change was. Most of the work was understanding the problem precisely enough that the fix became obvious, and that understanding took far longer than the fix itself.
| Approach | Best for | Watch out for |
|---|---|---|
| Baseline | Quick comparisons | No tail visibility |
| Instrumented | Finding the real hotspot | Overhead of the probes |
| Production | Truthful numbers | Harder to reproduce |
Lessons
The numbers below are from my own machine and my own workload. Treat them as a method, not a benchmark: the point is how the measurement was taken, not the exact figures it produced.
What surprised me most was how small the final change was. Most of the work was understanding the problem precisely enough that the fix became obvious, and that understanding took far longer than the fix itself.
A risky operation or a likely mistake.
Setting the scene
I want to be honest about the trade-offs here. The approach I settled on solves the problem I had, not every problem of this shape, and there is a clear point where you should reach for something more heavy-duty.
What surprised me most was how small the final change was. Most of the work was understanding the problem precisely enough that the fix became obvious, and that understanding took far longer than the fix itself.
[deploy]
strategy = "rolling"
healthcheck = "/healthz"
max_unavailable = 1The naive approach
By the end of this post you should be able to reproduce the whole thing yourself in under an hour, including the parts I got wrong the first time — those are marked so you can skip them if you're in a hurry.
I kept a log of every decision during this project, and reading it back a month later was humbling: the choices that aged well were the boring ones, and the clever ones all came with a footnote I hadn't written yet.
- Read the error message twice before changing anything
- Reproduce it with the smallest input you can
- Write the failing test first, even if it feels ceremonial
- Only then reach for the fix
What the profiler said
I want to be honest about the trade-offs here. The approach I settled on solves the problem I had, not every problem of this shape, and there is a clear point where you should reach for something more heavy-duty.
The first version worked, which is exactly why it was hard to see how wrong it was. Everything looked fine in the demo, and everything looked fine until traffic arrived on a Tuesday afternoon and the graph started doing something I'd never seen before.
The numbers below are from my own machine and my own workload. Treat them as a method, not a benchmark: the point is how the measurement was taken, not the exact figures it produced.
Wrap-up
If you remember one thing, make it this: a green build, a broken release, and the missing piece between them compounds. The cost of getting it right early is a afternoon; the cost of getting it late is a migration.